Running inference of UCE on multiple GPUs

[gmurtaza404]

Hello!

When I try to run inference with multiple GPUs, the code crashes with the error:

"Traceback (most recent call last):
File "/UCE/eval_single_anndata.py", line 155, in
main(args, accelerator)
File "/UCE/eval_single_anndata.py", line 85, in main
processor.run_evaluation()
File "/UCE/evaluate.py", line 145, in run_evaluation
run_eval(self.adata, self.name, self.pe_idx_path, self.chroms_path,
File "/UCE/evaluate.py", line 239, in run_eval
batch_sentences = model.module.pe_embedding(batch_sentences.long())
^^^^^^^^^^^^
File "***/UCE/env/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1695, in getattr
raise AttributeError(f"'{type(self).name}' object has no attribute '{name}'")
AttributeError: 'TransformerModel' object has no attribute 'module'. Did you mean: 'modules'?

Any suggestions on how to fix this?

[Yanay1]

Hi, please add the argument --multi_gpu=True and run the command with accelerate.